The invention generally relates to a laminate as well as a device and a method for manufacturing a laminate. More specifically, the invention relates to a laminate composed sequentially of an amorphous or polycrystalline substrate, a textured buffer layer, and an oriented oxide thin layer. The invention is also based on a device for manufacturing laminate with a vacuum chamber in which positioning devices for a substrate and buffer layer material dispensing devices for providing a buffer layer as a substrate for an oriented thin layer on the substrate are so arranged that buffer layer material is capable of being evaporated from the buffer layer material dispensing devices at an angle .alpha..sub.1.noteq.0 to the normal to the substrate surface, onto the latter, as well as a method for manufacturing laminate whereby a buffer layer is applied to a substrate, with the buffer layer material being evaporated from the buffer layer material dispensing devices at an angle .alpha..sub.1.noteq.0 to the normal to the substrate surface onto the latter, before an oriented thin layer is evaporated, as a way of manufacturing monocrystalline thin layers on polycrystalline or amorphous substrates.
To produce monocrystalline thin layers of a specific material, the material is generally applied to suitable monocrystalline substrates of another material. The substrates must have a suitable lattice structure in order to permit so-called heteroepitaxy, in which the monocrystalline structure of the substrate is assumed by the layer applied. This method is also used to produce thin layers of high-quality oxide high-temperature superconductors for example, such as YBa.sub.2 Cu.sub.3 O.sub.7-d (YBCO). In such superconductors, grain boundaries can drastically deteriorate the superconducting properties. Grain boundaries with a large grain boundary angle have a greater effect than those with a small angle between the crystal axes of the grains involved. The effect of the grain boundaries is obvious from a comparison of the critical current densities. This value is 3 to 5 MA/cm.sup.2 in YBCO at 77K on monocrystalline substrates in an intrinsic magnetic field. Typically, only 0.02 MA/cm.sup.2 is reached on nontextured substrates. For this reason, a monocrystalline structure of the superconductor is required when manufacturing high-quality, high-temperature superconductor layers.
The method of manufacturing high-quality superconductor layers by heteroepitaxial growth on monocrystalline substrates is limited to relatively small areas since such substrates are available only up to a very limited size. In addition, monocrystalline substrates are very expensive and therefore not economical in many cases. In particular it is not possible to make strips of monocrystals that are the prerequisite for power cables or wound magnets made of high-temperature superconductors.
Recently a number of approaches have been taken to circumvent the limitation to monocrystals. By using suitable methods, a quasi-monocrystalline or especially a biaxial texture is created in the substrate itself or in a buffer layer that is deposited on the substrate. This means that the crystal axes of the substrate or possibly of the buffer layer are aligned with a certain degree of unsharpness that is generally characterized by one or more half-width values. Then, in a second step for example, the superconductor is applied heteroepitaxially to this substrate or buffer layer. In this way, for example, improved superconducting properties are achieved, such as an increase in critical current density. This increases inversely with the half-width value or values of the buffer layer. Four important methods of producing a biaxial structure have been published.
Thus a method has been developed in which a biaxial texture is produced in nickel strips by multiple rolling followed by recrystallization (Rolling Assisted Biaxially Textured Substrates: RABiTS.RTM.), as published in A. Goyal et al., APL 69, Page 1795, 1996. However, a superconductor cannot be applied directly to these nickel strips since nickel does not have a substrate suitable for direct deposition of a superconductor due to diffusion and oxidation problems. Therefore a diffusion barrier is produced by a complex series of buffer layers grown epitaxially on it to produce a surface suitable for deposition of superconductors, on which surface the high-temperature superconductor layer then grows epitaxially. It is disadvantageous in this method that nickel is unsuitable for many applications because of its deficient tensile strength. In addition, the ferromagnetism of nickel is disadvantageous in applications involving a magnetic field.
In so-called "lon Beam Assisted Deposition" (IBAD) according to Y. likama et al., APL 60, page 769, 1992, the buffer layer is bombarded with low-energy ions as it is being deposited on a substrate, at a sharp angle. Y.sub.2 O.sub.3 -stabilized zirconium dioxide (YSZ) is usually employed as the buffer layer material. This method allows biaxially textured layers of high quality to be produced that permit deposition of superconducting films for example with very good properties. However the cost of the apparatus is high because an ion source is used, the deposition rate is low, and the deposition area is limited by the ion source. These points are cost-intensive and make the IBAD method unsuitable for commercial applications.
Another method is laser deposition at sharp angles (Inclined Substrate Deposition: ISD) as described in K. Hasegawa et al., Proc. of ICEC 16, 1996, Kitakyushu, Japan, and EP 669 411 A2. Depending on the deposition conditions, during laser deposition on a noninclined substrate, a crystal axis is set perpendicular to the substrate surface and hence parallel to the deposition direction. If the normal of the substrate is then inclined relative to the deposition direction, this crystal axis follows the deposition direction. At a suitable inclination angle, a second crystal axis is also oriented parallel to the surface and the biaxial texture is obtained. However, the laser deposition method is less suited for economical large-area coating because of the limited size of the area that can be coated at one time.
Similarly, a biaxial texture has been produced in thin metal layers by evaporating aluminum from resistance-heated boats (ct. T. Hashimoto et al., Thin Solid Films 182, 197, 1989). A background pressure of 4.times.10.sup.-5 mbar was used during deposition. The degree of texturing was poor, however. In addition, metals in general and aluminum in particular are not suitable as substrates for direct deposition of a superconductor because of diffusion and oxidation problems and a lack of thermal stability.
The goal of the present invention is to produce a laminate that is improved over the prior art and to improve devices and methods for manufacturing a laminate.